I conducted multiple data analysis projects by using R and Python, and focused on interpreting, engineering data and building models with the principles of machine learning. My projects response to the solutions for real-world concerns in resource allocation.
Bike share space-time prediction, Philadelphia
The algorithm model aims to predict Indego bike share demand in Philadelphia by taking time, physical and socio-economic factors into account. The model can assist Indego to better allocate bike resources based on demand varying across time and space.
EMS call prediction, Virginia Beach
With existing EMS call records data, this project predicted space and time distribution of future EMS calls by adopting Poisson Regression and Linear Regression, and an APP wireframe was also developed for ambulance drivers by providing the fastest route to direct to the right place based on modeling results.
Retail theft risk prediction, Chicago
This project predicts retail theft risk in Chicago based on crime records and various risk factors, and also figures out potential risk hotspots.
Larcerny index prediction, San Francisco
The Python project predicts larcerny risk in San Francisco by using “larceny index” as the dependent variable. It includes a variety of interactive visuals for exploratory analysis and modeling.
I also equipped myself with ArcGIS to work on spatial analysis with a focus on visualization trough map renderings: